Method and apparatus for skin color correction and digital photographing apparatus using both

1. A method of correcting skin color in an image, the method comprising:

extracting skin colors with a skin color extracting unit;

analyzing the extracted skin colors to calculate axes of a different vector space,

wherein the analyzing utilizes:

a principal component analyzing unit configured to perform a principal components analysis (PCA) using skin colors extracted by the skin color extracting unit to calculate axes for a principal component analysis vector space; and

a processor of an image processing device;

transforming the extracted skin colors to the PCA space;

using the result of the transformed skin colors to identify a color from an input image as a skin color; and

correcting the color identified as a skin color.

2. The method of claim 1 , further comprising:

detecting a skin region in the input image;

and wherein extracting further comprises:

extracting skin colors from the detected skin region.

3. The method of claim 1 , wherein extracting further comprises:

extracting skin colors from a skin training image.

4. The method of claim 2 , wherein detecting a skin region further comprises:

detecting a face region as a skin region in an input image.

5. The method of claim 1 , wherein analyzing further comprises:

analyzing the extracted skin colors to calculate axes of a different vector space, the axes being calculated using principal components analysis (PCA).

6. The method of claim 1 , wherein projecting further comprises:

projecting the transformed colors to fewer dimensions of the different vector space.

7. The method of claim 4 , wherein skin color regions are extracted from the face region.

8. The method of claim 6 , further comprising:

generating a lookup table according to the result of projecting the skin colors to fewer dimensions of the different vector space; and

wherein using the result further comprises: using the lookup table to identify a color from the input image as a skin color.

9. The method of claim 6 , wherein projecting further comprises:

projecting the extracted skin colors to fewer dimensions of the different vector space, the fewer dimensions excluding a principal axis, wherein the principal axis is the axis having the largest eigenvalue of the axes of the different space.

10. The method of claim 1 , wherein extracting further comprises:

extracting 3-dimensional skin colors; and

wherein analyzing further comprises:

analyzing the extracted skin colors to calculate axes of a 3-dimensional PCA vector space; and

projecting the extracted skin colors to 2-dimensions of the 3-dimensional PCA vector space.

11. The method of claim 8 , wherein the lookup table comprises skin colors projected to 2-dimensions of the different vector space.

12. The method of claim 1 , wherein correcting the color identified as a skin color further comprises:

extracting the color from the input image;

transforming the color to the new vector space; and

identifying the color as a skin color based on a closeness of the transformed color to the transformed skin colors in the new vector space,

wherein identifying the color as a skin color further comprises:

identifying the color as a skin color further based on identifying skin regions in the input image and identifying the color as a skin color based on the geographic closeness of the color to an identified skin region in the input image.

13. The method of claim 1 , wherein colors are represented using one of the following color models: red, green, and blue (RGB); hue-intensity-saturation (HIS); luminance and hue (YIQ); and, luminance and chrominance (YCC).

14. An apparatus for skin color correction, the apparatus comprising:

a face region detecting unit configured to detect a face region in an input image;

a skin color extracting unit configured to extract skin colors from the detected face region;

a principal component analyzing unit configured to perform a principal components analysis using skin colors extracted by the skin color extracting unit to calculate axes for a principal component analysis (PCA) vector space and transform the extracted skin colors to the PCA space; and

a skin color correcting unit configured to identify a color of the input image as a skin color using the result of the transformed skin colors to the PCA space and configured to correct a color identified as a skin color.

15. The apparatus of claim 14 , wherein the principal component analyzing unit is further configured to project the transformed extracted skin colors to 2-dimensions of the PCA space.

16. The apparatus of claim 15 , wherein the principal component analyzing unit is further configured to project the transformed extracted skin colors to 2-dimensions of the PCA space, the 2-dimensions not including a principal axis, wherein the principal axis is an axis having the largest eigenvalue from among the axes of the PCA space.

17. The apparatus of claim 14 , further comprising a lookup table generating unit for generating a lookup table according to the result of the transformed skin colors to the PCA space,

wherein the skin color correcting unit identifies a color as a skin color using the lookup table.

18. The apparatus of claim 14 , wherein the skin color correcting unit comprises:

a skin region separating unit for separating skin regions from the input image according to skin color;

a skin correction region designating unit for designating regions to perform skin color correction from among the separated skin regions; and

a skin color correction processing unit for performing a skin color correction on the designated regions.

19. A non-transitory computer readable recording medium encoded with a computer executable program to perform the method according to claim 1 .